516 Additional verification of patient positioning in fractionated stereotatic radiotherapy using electronic portal imaging

$220 Table 1. Level of axial CT slice

"-" - caudal; "+" - cranial

-2.0 cm -1.5 cm -1.0 cm - 0.5 cm symphysJs pubis - 0 cm 0.5 cm 1.0 cm 1.5 cm 2.0 cm

Posters Mean distance from anterior border of the symphysis pubis to the CTV-rectal interface (cm) l-large 2 3 4 rectum, medium medium medium rectum, rectum, large rectum, bladder medium small large bladder bladder bladder 5.1 5.6 5.8 6.1 5.7 5.6 5.1 4.9 5

5.2 5.8 6.5 7.4 7.6 7.8 7 6.7 6.5

5.3 5.9 6.5 7.9 8.2 8.1 7.6 7.3 7.2

5 6 6.2 7 7 7.4 6.9 7 6.8

Conclusions: Variations in rectal and bladder filling have a major impact on movement of the CTV - rectal interface which is a critical region for definition of prostatic fossa target definition, particularly above the symphisis pubis. This study suggests that for most patients a PTV of 1 cm is adequate however for those with a high risk of recurrence outside the true pelvis in the region of the seminal vesicle or retrovesicle space a planning target extension of at least 1.5cm is recommended to account for variations in target position.

515 Systematic error reduction using a novel patient set-up correction strategy, S. M~nsson, H. NystrOm, M. Bj6rk Copenhagen University Hospital, Department of Radiooncology, Copenhagen, Denmark Purpose: The relation between the systematic (Zs~t-.p) and random (crset_up) errors vary for different kind of treatments. The purpose of this study is to find an optimal patient set-up correction strategy for various ~set_up/O-set-up relations and to apply the strategy to measured clinical data. Material and Methods: A computer program (MATLAB) simulated set-up deviations for 1000 "patients" to investigate the efficiency of the Shrinking Action Level (SAL) strategy, No Action Level (NAL) strategy, the Shalev's strategy, the current strategy used at the department, and a novel strategy called the M3M strategy. To verify the simulation result, a clinical study was carried out. Using EPID, 765 portal images from 17 patients were acquired. All patients had the same set-up procedure and were imaged at each fraction. Results: The computer simulation study showed that for the investigated relations Zset-upS = O'set-upS J 2 and Z~set-upS = CTset,ps, the optimal correction strategy is the M3M strategy. In the Zset-,pS = ¢,et-,ps / 2 case, all the other correction strategies actually have no effect or even decrease the accuracy, i.e. the mean deviation after correction is increased. Even when SAL and Shalev's strategies are most effective within the limits of this simulation, i.e. when Zset-upS / ~,et-.ps = 3, the M3M strategy still reduces the mean deviation a further 2.5 times. The systematic error, Zset up, was found to be approximately equal to the random error, cruet_up for the pelvic patients investigated. Applying the M3M strategy to the measured clinical data, the Zset-,p decreased in the posterior-anterior field from 3.1 to 1.5 mm in the lateral direction and from 2.2 to 0.7 mm in the caudal-cranial direction. In the lateral field, Zset-up decreased from 3.3 to 1.3 mm in the dorsal-ventral direction and from 1.6 to 0.5 mm in the caudal-cranial direction. Conclusions: The computer simulated study shows that the M3M strategy is the most successful among all the investigated correction strategies and for the most common zset-up /crset-,p - relations. Unlike other correction strategies M3M always reduces the systematic deviation. Hence using

the M3M strategy can reduction.

potentially contribute

to

margin

516 Additional verification of patient positioning in fractionated stereotatic radiotherapy using electronic portal imaging V. Batel, S. Germano, P. Pereira, F. Marques, F. Mascarenhas Santa Maria Hospital, Radiotherapy Department, Lisbon, Portugal Fractionated stereotactic radiotherapy (FSRT) offers a technique to minimize the absorbed dose to normal tissues; therefore, quality assurance is essential for these procedures. At our department FSRT of small brain tumours is performed using the GilI-Thomas-Cosman (GTC) relocatable stereotactic frame. Recently, we introduce a additional verification of patient positioning in this treatments. Before the first day of treatment the patient is CT scanned to confirm the correct local/sat/on of the isocenter. Daily positioning verification is performed with a depth helmet and measuring probe. The GTC frame is refitted when differences from planning CT are worse then 2 mm for two or more measuring portals. In recent times, we initiated an additional verification of patient positioning based on electronic portal imaging. Two images (one anterior-posterior and one lateral) per patient are acquired on the first day of the treatment and repeated approximately two weeks after, depending on the total dose and fractionation. The determination of the displacement vector from portal images is not easy. Although we verified deviations in the same direction as with the depth helmet. A maximum value of 3 mm was measured. The measurements with depth helmet are a good method to correctly position the frame on the patient. Portal images, as an additional verification, do not account for small deviations less then 1 mm.

Miscellaneous 517 A conformal technique for breast irradiation and minimal lung involvement A. Foqliata, A. N/colin/, L. Cozzi Oncology Institute of southern Switzerland, Medical Physics, Bellinzona, Switzerland Objective: To appraise the benefit of a conformal technique achieve high conformal avoidance of lung tissue in the treatment of breast. Material and methods: A comparative study carried out at planning level was designed for patients selected for their inadequate sparing of healthy lung tissue with the tangential technique. A threshold of about 30% of the prescribed dose for the mean lung dose (MLD) was set as tentative inclusion criteria. Dose plans were designed for the conventional (reference) tangential technique, for an alternative conformal approach based on three isocentric beams and for the newly proposed technique consisting of two non coplanar beams. Planning objectives were set to minimise V20Gy and MLD. Results: In average for the new technique compared to the reference, MLD dropped from 31.6% to 21%, V20Gy from 29.5% to 18.2% and more strikingly, the dose delivered to 1/3 (1/4) of the lung volume dropped from 28.5% (67.3%) to 8.7% (13.4%). For the target volume, the volume receiving at least 90% (95%) of the prescription resulted 97.4% (87.7%) for the new and 97.3% (84.3%) for the reference. Equivalent uniform dose resulted 48.5 Gy for the new and 49.1 Gy for the reference (for a prescription of 50 Gy at 2 Gy/fraction). Conformity index improved significantly from 2.58 for the reference to 1.84 for the new technique. Conclusion: For a subgroup population of breast cancer patients, a treatment modality with non coplanar beams was